A transmission circuit for transmitting data of varying priorities on a network medium is provided. The transmission circuit includes sub-circuits to receive and store data frames into random access memory frame buffers and priority tables. Sub-circuit priority resolution selects the highest priority frame, and sub-circuit frame transmission transmits the frame to a media access controller to be made available by the network medium.
|
9. A method of transmitting a highest priority data frame available for transmission in a frame buffer, the method comprising:
a) reading data from a register to determine a priority of the highest priority data frame available for transmission;
b) locating an address at which the highest priority frame is stored in a frame buffer;
c) writing the address of the highest priority data frame to a frame transmission circuit;
d) overwriting the address of the highest priority data frame with an address of a new highest priority data frame if a new higher yet priority data frame becomes available; and
e) retrieving the new highest priority data frame from the frame buffer and transmitting the new highest priority data frame when the network media is available.
1. A frame processing unit for transmitting data frames of varying priorities on a network medium comprising:
a) a frame buffer management circuit receiving data frames and storing data frames in a buffer memory;
b) a register storing data representing the existence of data frames of a designated priority in the buffer memory;
c) a priority resolution circuit, reading the register to determine a highest priority data frame available for transmission; and
d) a frame transmission circuit receiving an address of the highest priority data frame from the priority resolution circuit, receiving a signal from a media access controller indicating that the data frame may be transmitted, retrieving the data frame from the buffer memory corresponding to the address, and making the data frame available to the media access controller for transmitting to the network medium,
wherein the priority resolution circuit continually retrieves data from the register to determine a highest priority data frame in the buffer memory and replaces the address previously provided to the frame transmission circuit if a higher priority frame becomes available.
13. A network computer comprising:
a) a central processing unit operating a plurality of applications generating data frames of varying priorities for transmission on a network medium;
b) a network interface circuit receiving the data frames and transmitting the data frames on the network medium in priority order, the network interface circuit including:
i) a frame buffer management circuit receiving data frames from the central processing unit and storing data frames in a buffer memory;
ii) a register storing data representing the existence of data frames of a designated priority in the buffer memory;
iii) a priority resolution circuit, reading the register to determine a highest priority data frame available for transmission; and
iii) a frame transmission circuit receiving an address of the highest priority data frame from the priority resolution circuit, receiving a signal from a media access controller indicating that the data frame may be transmitted, retrieving the data frame from the buffer memory corresponding to the address, and making the data frame available to the media access controller for transmitting to the network medium,
wherein the priority resolution circuit continually retrieves data from the register to determine highest priority data frame in the buffer memory and replaces an address previously provided to the frame transmission circuit if a higher priority frame becomes available.
2. The frame processing unit of
3. The frame processing unit of
4. The frame processing unit of
5. The frame processing unit of
6. The frame processing unit of
7. The frame processing unit of
8. The frame processing unit of
10. The method of
11. The method of
12. The method of
14. The network computer of
15. The network computer of
16. The network computer of
17. The network computer of
18. The network computer of
19. The network computer of
20. The network computer of
|
The present invention relates generally to network interfacing, and more particularly, to an apparatus and method for prioritizing data frames for transmission on a network medium.
As computer engineering and digital signal processing technology has advanced, there has been an increasing demand for cost-efficient transmission of digital information through communication networks. To meet this demand, high-speed packet-switched communication networks have been developed. The packet-switched communication network typically multiplexes different information sources into a single communication channel to maximize bandwidth utilization. For example, in a packet-switched network, computer data files, digitized voice data, and other data content are coded into transmission frames. Each data frame transmission is then transmitted to a remote device on a network medium when the channel is available.
A problem with such networks is that during peak transmission periods, the network can become congested. When the network is congested, data frames are held in queues of transmitters and switching nodes, causing delays in delivery of data frames. Traditionally, data frames are transmitted in the order they are received, first-in-first-out (FIFO), irrespective of a data frame's priority.
When data frames containing computer data files or other computer data content are delayed, the delay may be noticeable and annoying to a user waiting for a file or a web page to load. However, when the file arrives, it is just as useful and provides the same information content to the computer or the user as if it had arrived in a faster time. This can be referred to as non-time sensitive data or non-real time data, i.e., lower priority data frames.
On the other hand, data frames that contain digitized voice data representing voice communication, such a telephone call between two operators, are time sensitive or real time data, i.e., higher priority data frames. When speech is digitized, segmented, and compressed into speech frames, each data frame must arrive at the receiver within a fixed time window for the receiver to decompress and reconstruct to an analog audio signal. Network delay of time sensitive packets, such as digitized voice data, will result a broken audio signal at the receiver and/or completely unintelligible sound bursts. In either case, time sensitive data, unlike non-time sensitive data, is useless if it does not arrive on time because of congested networks. While digital audio data and digital video data are obvious examples of time sensitive data, other types of data in any transaction processing system can have varying priority requirements for network resources.
One solution to relieve network congestion and to ensure timely delivery of all data frames regardless of time sensitivity or priority is simply to increase overall network bandwidth by increasing the data rate and/or adding additional transmission lines and/or routers. However, such solution can be costly, and the additional resources are not needed during periods the network is not congested.
Another solution to ensure timely deliver of time sensitive data frames is to prioritize data frames within the queue. However, prioritizing frames within a queue does not resolve a front of line blocking problem. A front of line blocking problem occurs when, for example, the highest priority data frame (say priority 3) is retrieved from a queue and is written to a register (or other memory) for transmission in the next available time slot (e.g. interval of time available to the media access controller for transmission). At this time, that data frame is isolated from the remaining data frames left in the queue. The remaining data frames in the queue may be reprioritized with newer, incoming data frames, however, no other frames can be transmitted until that first data frame is transmitted. Hence, a higher priority data frame (say 6) which has come into the queue after the first frame was written to the register is blocked from transmitting before the lower priority data frame 3.
What is needed is a transmitter system and a method that provides for higher priority frames to be prioritized over lower priority frames which does not suffer the front of line blocking problems associated with known prioritization systems.
A first aspect of the present invention is to provide a frame processing unit for transmitting data frames of varying priorities on a network medium. The frame processing unit comprises: a) a frame buffer management circuit receiving data frames and storing data frames in a buffer memory; b) a register storing data representing the existence of data frames of a designated priority in the buffer memory; c) a priority resolution circuit, reading the register to determine the highest priority data frame available for transmission; and d) a frame transmission circuit receiving an address of the highest priority data frame from the priority resolution circuit, receiving a signal from a media access controller indicating that a frame may be transmitted, retrieving a frame from the buffer memory corresponding to the address, and making the data frame available to the media access controller for transmitting to the network medium.
Further, the priority resolution circuit may continually retrieve data from the register to determine the highest priority data frame in the buffer memory and replace an address previously provided to the frame transmission circuit if a higher priority frame becomes available.
The frame buffer may be a random access memory frame buffer and the processing unit may further include a random access memory pointer table storing an indicator of the priority for each frame in the frame buffer along with the address location of each frame in the fame buffer. The frame buffer management circuit may locate the address of the highest priority frame, as indicated by the register, from looking up the priority in the random access memory pointer table.
In one embodiment, the media access controller receives the frame from the frame transmission circuit and makes each frame available to physical layer circuitry. Thereafter, the frame transmission circuit may send a command to the priority resolution circuit which in turn updates the register and the random access memory pointer table to reflect transmission of the frame.
The frame buffer management circuit may receive and store data frames from an application via a peripheral bus and the data received via the peripheral bus may include data of varying priorities as assigned by the application.
A second aspect of the present invention is to provide a method of transmitting the highest priority data frame available in a frame buffer. The method comprises: a) reading data from a register to determine the priority of the highest priority data frame available for transmission; b) locating a frame buffer address at which the highest priority frame is stored in a frame buffer; c) writing the address of the highest priority data frame to a frame transmission circuit; d) overwriting the address of the highest priority data frame with the address of a new highest priority data frame if a new higher yet priority data frame becomes available; and e) retrieving the new highest priority data frame from the frame buffer and transmitting the new highest priority data frame when the network media is available.
The step of locating the frame buffer address may include looking up the frame buffer address in a pointer table which stores the frame buffer address along with the priority of the frame stored at the address. Further, the method may further include updating the register and updating the pointer table upon transmission of a data frame to reflect transmission of the data frame.
A third aspect of the present invention is to provide a network computer comprising a central processing unit operating a plurality of applications generating data frames of varying priorities for transmission on a network medium. The network computer includes a network interface circuit receiving the data frames from the central processing unit and transmitting the data frames on the network medium in priority order. The network interface circuit includes: a) a frame buffer management circuit receiving data frames from the central processing unit and storing data frames in a buffer memory; b) a register storing data representing the existence of data frames of a designated priority in the buffer memory; c) a priority resolution circuit, reading the register to determine the highest priority data frame available for transmission; and d) frame transmission circuit receiving an address of the highest priority data frame from the priority resolution circuit, receiving a signal from a media access controller indicating that a frame may be transmitted, retrieving a frame from the buffer memory corresponding to the address, and making the data frame available to the media access controller for transmitting to the network medium.
Further, the priority resolution circuit may continually retrieve data from the register to determine the highest priority data frame in the buffer memory and replace an address previously provided to the frame transmission circuit if a higher priority frame becomes available.
The frame buffer may be a random access memory frame buffer and the processing unit may further include a random access memory pointer table storing an indicator of the priority for each frame in the frame buffer along with the address location of each frame in the fame buffer. The frame buffer management circuit may locate the address of the highest priority frame, as indicated by the register, from looking up the priority in the random access memory pointer table.
In one embodiment, the media access controller receives the frame from the frame transmission circuit and makes each frame available to physical layer circuitry. Thereafter, the frame transmission circuit may send a command to the priority resolution circuit which in turn updates the register and the random access memory pointer table to reflect transmission of the frame.
The frame buffer management circuit may receive and store data frames from an application via a peripheral bus and the data received via the peripheral bus may include data of varying priorities as assigned by the application.
The present invention will now be described in detail with reference to the drawings. In the drawings, like reference numerals are used to refer to like elements throughout.
Referring to
Coupled to network 10 is a plurality of client workstations 18(a)-18(d) which, in the preferred embodiment, are typical desktop computers. Each client workstation 18 includes appropriate hardware and software for communicating over a data network. For example, each workstation 18 may be HPNA 2.0 enabled and the network medium 14 may be a POTS twisted pair telephone network. HPNA 2.0 is a protocol for transferring data over POTS twisted pair telephone wiring that is promulgated by the Home Phone line Networking Association which is an industry consortium including Advanced Micro Devices of Sunnyvale Calif. Further, each client workstation 18 operates a data processing application which interfaces with an application server 16 via the network 10. Additionally, each client workstation 18 is H.323 enabled for enabling the operator to carry on full duplex audio communications (e.g. telephone calls) with other operators of workstations 18 and other people using the data network 10 via an H.323 telephony gateway 20. H.323 is an Internet protocol (IP) telephony standard promulgated by the International Telephony Union (ITU).
Application server 16 is a typical application server storing and communicating files with each of the client workstations 18. Telephony gateway 20 functions to interconnect digitized audio data frames (e.g. frames of digitized audio data representing telephone calls) between multiple client workstations 18 and/or standard telephones coupled to a PBX system or a local telephone company subscriber loop.
Referring to
The client workstation 18 includes a typical keyboard 30 and display 32 through which an operator interfaces with the data processing application 60. The keyboard 30 is coupled to keyboard interface circuitry 46, which in turn couples to the processor 42 via peripheral bus 44. A keyboard driver 56 stored in memory 48 drives keyboard 30 using known techniques. Similarly, display 32 is coupled to display interface circuitry 47, such as a video card, which in turn couples to processor 42 via peripheral bus 44. A display driver 54 stored in memory 48 drives the display interface circuitry 47 and display 32.
A LAN telephony system 34 enables the workstation operator to initiate and receive telephone calls via the H.323 telephony gateway 20 (FIG. 1). LAN telephony system 34 includes a speaker 36, a microphone 38, and an audio subsystem 40. Audio subsystem 40 couples to the processor 42 via peripheral bus 44. An audio subsystem driver 58 stored in memory 48 operates the audio subsystem 40, speaker 36, and microphone 38 for audio interface with the operator. An audio communication application 50 functions to encode/decode frames of digital audio data in accordance with the H.323 standard as well as exchange such digital audio data frames with telephony gateway 20 (FIG. 1). The audio communication application 50 also enables the operator to use the keyboard 30 and display 32 as an interface for dialing or otherwise initiating a telephone call.
Client workstation 18 is coupled to the network medium 14 (i.e., 14(a), 14(b) or 14(c)) through a network interface card 62. The network interface card 62 is coupled to the processor 42 via peripheral bus 44 and a network interface driver 52 stored in memory 48 and executed by the processor 42 drives the network interface card 62.
It should be appreciated that network interface card 62 functions to communicate frames of digital audio data with telephony gateway 20 and frames of data processing application data with application server 16. As discussed above, when the network 10 or any of the sub networks 13 become congested, delivery of frames may be delayed. Further, delivery of frame to the telephone gateway 20 and to the application server 16 may be delayed due to heavy loading on the gateway 20 or the application server 16 (e.g. other workstations 18 trying to send frames simultaneously). While delay of frames containing data processing application data may be noticeable and annoying to a user waiting for a file or web page to load, the delays do not destroy the usefulness of the data. However, delays of digitized audio data representing voice communication or real time video data can result in a broken audio signal at the receiver and/or completely unintelligible sound bursts or the disruption of the video images. Thus, data frames of varying priorities containing digitized audio data or real time video data can be referred to as real time frames, i.e., data frames which have been assigned higher priorities by an application. Data frames of varying priorities containing data processing application data can be referred to as non-real time frames, i.e., data frames which have been assigned lower priorities by an application.
Referring to
In operation, the frame buffer management circuit 100 will function to read incoming data frames received from peripheral bus 44 and to write the data frames to the random access memory frame buffer 102. Further, the frame buffer management circuit 100 writes, to the pointer table 104, the start address and end address corresponding to where the data frame was stored in the frame buffer 102 along with the corresponding priority level. The frame buffer management circuit 100 also sets a bit in the register 106 corresponding to the priority level.
In operation, the priority resolution circuit 108 will function to read the register 106 to determine the highest priority data frame available for transmission, to retrieve the frame's address from the random access memory pointer table 104, and to send the data frame's address to the frame transmission circuit 110. Further, after a frame has been transmitted, the priority resolution circuit clears the random access memory pointer table 104 and, if appropriate, the indicator in the register 106.
In operation, the frame transmission circuit 110 will function to retrieve the data frame from the random access memory frame buffer 102 and to present the data frame to a media access controller 72. Generally real time data frames will have a high priority level indicator while non-real time data frames will have a low priority indicator. A more detailed discussion of the operation of processing unit 74 is included later herein with respect to
In operation, the media access controller 72 is coupled to a physical layer circuit 68. The physical layer circuit 68 includes digital signal processing circuits for payload encoding bits of data within the transmission data frame and generating a digitized modulated carrier representing the transmission data frame. A digital to analog converter 70 generates an analog carrier signal on line 71. An analog front end 66 couples the analog carrier signal on line 71 to the network medium 14 and includes appropriate amplifier circuits for assuring that the strength of the signal is within the parameters of the network transmission protocol.
In operation, the media access controller 72 receives a signal from channel sensor circuitry (not shown) on line 73 indicating that the network medium 14 is available for transmission. Upon receipt of such signal, the media access controller 72 generates a data frame request to the frame transmission circuit 110. Generally, the frame transmission circuit 110 provides the highest priority data frame stored in the random access memory frame buffer 102 to the media access controller 72 for transmission. However, it is contemplated that in certain environments, the media access controller 72 may request a specific priority frame that is less than the highest priority frame. In such environment, the frame transmission circuit 110 will provide the requested priority framed to the media access controller 72.
At step 80, the frame buffer management circuit 100 monitors the peripheral bus 44 to determine whether a data frame is present for transmission. If a data frame is not available, as indicated by return loop 81, the frame buffer management circuit 100 waits until a data frame is available from the peripheral bus 44. At step 82, if a data frame is present on the peripheral bus 44, the frame buffer management circuit 100 writes the data to the random access memory frame buffer 102. At step 83, the frame buffer management circuit 100 writes the data frame's address and the frame's priority level to the random access memory pointer table 104. And, at step 84, the frame buffer management circuit 100 sets a bit in the register 106 corresponding to the data frame's priority level in the random access memory pointer table 104.
Referring to
At step 89, the priority resolution circuit 108 waits for a signal from the frame transmission circuit 110 to clear the register 106 and the random access memory frame buffer 102 and the random access memory pointer table 104. At step 90, if the priority resolution circuit 108 receives a signal to clear the register 106 from the frame transmission circuit 110, the priority resolution circuit 108 clears the register 106. At step 91, the priority resolution circuit 108 clears the random access memory frame buffer 102 and the random access memory pointer table 104. If at step 89, the priority resolution circuit 108 does not receive a signal from frame transmission circuit to clear the register 106, the priority resolution circuit 108 rereads the register 106 for determining whether a higher priority data frame (or a frame with the priority requested by the media access controller) is available as represented by step 92. If there is no such frame, as indicated by return loop 93, the priority resolution circuit 108 again waits for a signal from the frame transmission circuit 110 to clear the register 106 (step 89). At step 94, if there is such a data frame available, the priority resolution circuit 108 retrieves the address of such data frame from the random access memory pointer table 104. At step 95, the priority resolution circuit 108 writes the data frame's address to the frame transmission circuit 110.
It should be appreciated that steps 89 and 92 operate to assure that any address written to the frame transmission circuit 110 can be overwritten with an address of a higher priority frame (or by a frame of the requested priority) at any time prior to the priority resolution circuit 108 receiving the clear signal at step 89.
Referring to
Referring to
The above described systems and methods provide a frame buffer for the prioritization of real time data frames for transmission over a network medium. Such prioritization may be independent of any prioritization scheme, if any, implemented in a media access controller.
The preferred prioritization scheme provides for eight priority levels, which can be represented by a three bit priority indicator. However, it should be appreciated that additional priority levels can be assigned to each frame by scaling the teaching of this preferred embodiment. A system with multiple priority levels becomes useful for prioritizing between real time audio data and real time video data for example. Such a system also enables prioritization between different digitized audio data frames to provide a higher priority for frames that contain more critical speech sounds. For example, frames that contain vowel sounds critical for operators understanding a phrase of speech may be prioritized over frames containing hard consonant sounds which, if dropped, may not render the speech completely unintelligible.
Although the invention has been shown and described with respect to certain preferred embodiments, it is obvious that equivalents and modifications will occur to others skilled in the art upon the reading and understanding of the specification. The present invention includes all such equivalents and modifications, and is limited only by the scope of the following claims.
Patent | Priority | Assignee | Title |
7162419, | May 04 2001 | RPX Corporation | Method in the decompression of an audio signal |
7293103, | Feb 20 2001 | AT&T Corporation | Enhanced channel access mechanisms for a HPNA network |
7298757, | Feb 20 2001 | AT&T Corporation | Enhanced channel access mechanisms for an HPNA network |
7310326, | Feb 20 2001 | AT&T Corporation | Enhanced channel access mechanisms for an HPNA network |
7606258, | Feb 19 2002 | AT&T Intellectual Property II, L.P. | Enhanced channel access mechanisms for an HPNA network |
7606930, | Feb 20 2001 | AT&T Intellectual Property II, L.P. | Enhanced channel access mechanisms for an HPNA network |
7610399, | Feb 20 2001 | AT&T Intellectual Property, II, L.P. | Enhanced channel access mechanisms for an HPNA network |
7613823, | Feb 20 2001 | AT&T Intellectual Property II, L.P. | Enhanced channel access mechanisms for an HPNA network |
7707342, | Aug 31 2006 | SOCIONEXT INC | Method and apparatus for arbitrating access |
7779150, | Feb 20 2001 | AT&T Intellectual Property II, L.P. | Enhanced channel access mechanisms for an HPNA network |
8046414, | Oct 21 2003 | General Motors LLC | Method for accessing email attachments from a mobile vehicle |
9288157, | Oct 15 2013 | National Instruments Corporation | Time-sensitive switch for scheduled data egress |
Patent | Priority | Assignee | Title |
4719620, | Jan 11 1985 | Sharp Kabushiki Kaisha | Signal transmission system in a bus-type network |
4914650, | Dec 06 1988 | Lucent Technologies, INC | Bandwidth allocation and congestion control scheme for an integrated voice and data network |
5043981, | May 29 1990 | RPX Corporation | Method of and system for transferring multiple priority queues into multiple logical FIFOs using a single physical FIFO |
5077655, | Aug 20 1987 | Fujitsu Limited | Communication control device in computer system |
5381413, | Dec 28 1992 | CONGRESS FINANCIAL CORPORATION NEW ENGLAND , A MASSACHUSETTS CORPORATION | Data throttling system for a communications network |
5434976, | Sep 28 1992 | Microchip Technology Incorporated | Communications controller utilizing an external buffer memory with plural channels between a host and network interface operating independently for transferring packets between protocol layers |
5448564, | Jan 31 1994 | Advanced Micro Devices, Inc. | Modular architecture for fast-packet network |
5671355, | Jun 26 1992 | PREDACOMM, INC | Reconfigurable network interface apparatus and method |
5926458, | Jan 31 1997 | AVAYA Inc | Method and apparatus for servicing multiple queues |
6097734, | Apr 30 1997 | RPX Corporation | Programmable reassembly of data received in an ATM network |
6157623, | Feb 14 1997 | GLOBALFOUNDRIES Inc | Apparatus and method for selectively outputting data using a MAC layer interface or a PCI bus interface |
6393457, | Jul 13 1998 | International Business Machines Corporation | Architecture and apparatus for implementing 100 Mbps and GBPS Ethernet adapters |
6442631, | May 07 1999 | Hewlett Packard Enterprise Development LP | Allocating system resources based upon priority |
Executed on | Assignor | Assignee | Conveyance | Frame | Reel | Doc |
Aug 17 2000 | GARG, ATUL | Advanced Micro Devices, INC | ASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS | 011030 | /0968 | |
Aug 17 2000 | ACHARYA, YATIN | Advanced Micro Devices, INC | ASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS | 011030 | /0968 | |
Aug 23 2000 | Advanced Micro Devices, Inc. | (assignment on the face of the patent) | / | |||
Jun 30 2009 | Advanced Micro Devices, INC | GLOBALFOUNDRIES Inc | AFFIRMATION OF PATENT ASSIGNMENT | 023119 | /0083 | |
Nov 17 2020 | WILMINGTON TRUST, NATIONAL ASSOCIATION | GLOBALFOUNDRIES U S INC | RELEASE BY SECURED PARTY SEE DOCUMENT FOR DETAILS | 056987 | /0001 |
Date | Maintenance Fee Events |
Dec 10 2004 | ASPN: Payor Number Assigned. |
Jun 30 2008 | M1551: Payment of Maintenance Fee, 4th Year, Large Entity. |
Jul 03 2008 | R1551: Refund - Payment of Maintenance Fee, 4th Year, Large Entity. |
Aug 01 2012 | M1552: Payment of Maintenance Fee, 8th Year, Large Entity. |
Oct 07 2016 | REM: Maintenance Fee Reminder Mailed. |
Mar 01 2017 | EXP: Patent Expired for Failure to Pay Maintenance Fees. |
Date | Maintenance Schedule |
Mar 01 2008 | 4 years fee payment window open |
Sep 01 2008 | 6 months grace period start (w surcharge) |
Mar 01 2009 | patent expiry (for year 4) |
Mar 01 2011 | 2 years to revive unintentionally abandoned end. (for year 4) |
Mar 01 2012 | 8 years fee payment window open |
Sep 01 2012 | 6 months grace period start (w surcharge) |
Mar 01 2013 | patent expiry (for year 8) |
Mar 01 2015 | 2 years to revive unintentionally abandoned end. (for year 8) |
Mar 01 2016 | 12 years fee payment window open |
Sep 01 2016 | 6 months grace period start (w surcharge) |
Mar 01 2017 | patent expiry (for year 12) |
Mar 01 2019 | 2 years to revive unintentionally abandoned end. (for year 12) |